Is Burst Hydrophobic Collapse Necessary for ... - ACS Publications

we find that folding to the native conformation is preceded by a rapid collapse into a burst ... Instead, we find an all-or-none transition directly f...
0 downloads 0 Views 4MB Size
3066

Biochemistry 1 9 9 5 , 3 4 , 3066-3076

Is Burst Hydrophobic Collapse Necessary for Protein Folding?? A. M. Gutin, V. I. Abkevich, and E. I. Shakhnovich* Department of Chemistry, Haward University, I 2 Oxford Street, Cambridge, Massachusetts 02138 Received August 25, 1994; Revised Manuscript Received November 30, 1994@

Folding of the lattice model of proteins is studied using Monte Carlo simulation. The amino acid sequence is designed to have a pronounced energy minimum for a given target (native) conformation. Our simulations reveal two possible scenarios. When the overall attraction between residues dominates, we find that folding to the native conformation is preceded by a rapid collapse into a burst intermediate which is a compact but .structureless globule. Then, after a much longer time, an all-or-none transition from the globule to the native conformation occurs. In contrast, when the overall attraction is not strong, we do not observe a burst collapse stage. Instead, we find an all-or-none transition directly from the coil to the native conformation. Both scenarios yield comparable rates of folding. On the basis of these findings we discuss the role of intermediates in thermodynamics and kinetics of protein folding. ABSTRACT:

The famous Levinthal argument about the impossibility of finding a protein’s structure by random search (Levinthal, 1969) stimulated the development of many phenomenological models aimed at explaining how a protein can fold, avoiding a random search over its huge conformational space (Karplus & Weaver, 1979; Weaver & Karplus, 1994; Kim & Baldwin, 1982; Harrison & Durbin, 1985; Wetlaufer, 1973). The underlying idea for most phenomenological models of folding concerned the existence of some pathway, different in different models, with several stages and intermediates. This introduced hierarchical elements to folding in which certain degrees of freedom become “frozen” so that the remaining search is feasible, since it required scanning over a much smaller number of conformations. A well-known example of this kind is a secondary structure framework or the related, diffusion-collision model (Kim & Baldwin, 1982; Karplus & Weaver, 1979; Weaver & Karplus, 1994). These models postulate that certain elements of native structure, microdomains (usually associated with elements of native secondary structure), are formed very fast, and in subsequent folding they move as a whole and associate via, e.g., the diffusioncollision mechanism. A similar way to decrease the effective number of conformations searched is to use a two-stage kinetics, the first stage being compactization into structureless globule. At the second stage the search for the native conformation takes place among only compact globular conformations whose number is much less than the total number of conformations (Orland et al., 1985; Dill, 1985). Therefore, compactization should facilitate the search for the native conformation. The compact state without specific structure (judged by lack of specific protection against H-D exchange in most parts of a protein) was indeed observed as a “burst” intermediate in a number of proteins: hen egg lysozyme (Redford et al., 1992; Itzhaki et al., 1994), cytochrome c (Chaffotte et al., 1991), and a tryptophan-containing mutant Supported by the Packard Foundation.

* Address correspondence to this author (telephone 6 17-495-4130;

FAX 617-496-5948; Internet: [email protected]). Abstract published in Advance ACSAbstructs, February 15, 1995. @

of ubiquitin at 25 “C (Khorasanizadeh et al., 1993). This burst intermediate is a compact conformation with few specific contacts and some secondary structure detected by CD which is likely to be fluctuating without any specific localization. The evidence for that is absence of specific protection of the amides in the burst intermediate. The two-stage kinetics was also observed in computer simulations of protein folding (Shakhnovich et al., 1991; Miller et al., 1992; Honeycutt & Thirumalai, 1992; Camacho & Thirumalai, 1993; Fukugita et al., 1993; Sali et al., 1994a; Socci & Onuchic, 1994). It was argued that burst compactization plays an important role in making the subsequent random search for a transition state feasible. However, it was noted (Sali et al., 1994) that such mechanism is likely to be applicable to folding of relatively short, prebiological peptides with almost random sequences. As chain length N increases to realistic sizes, random sequences fail to fold (Shakhnovich, 1994). This is due to the fact that for longer sequences the number of compact conformations is still exponentially large in N . The point of view that Levinthal paradox can be resolved only by sequential mechanism involving intermediates enjoyed indirect support from experiment since in many studies compact intermediates were found indeed. However, the generality of intermediates was questioned in a recent study of Sosnick et al. (1994), who showed that folding of cyt c at pH 5 differs dramatically from the scenario described in earlier studies and represents essentially the two-state kinetics where all observed properties become native-like simultaneously and fast. And cyt c is not unique. Another protein, chymotrypsin inhibitor 2, folds kinetically via a twostate mechanism showing no intermediates (Jackson & Fersht, 1991). Moreover, in a detailed study of ubiquitin folding, Khorasanizadeh et al. (1993) showed that, depending on the final concentration of the denaturing agent after dilution, the same protein may or may not have folding intermediates. Sosnick et al. (1994) suggested to distinguish “artificial or adventitious” barriers for folding (such as slow heme ligation in cyt c in their case) from the barriers intrinsic for folding. In the same way, one can question the role of folding intermediates because indeed there is a number of experimental examples where folding apparently proceeds

0006-2960/95/0434-3066$09.00/0 0 1995 American Chemical Society

Is Burst Hydrophobic Collapse Necessary without them. Moreover, one can imagine that burst compactization may result in slowing of certain chain motions because parts of the chain move in dense environment. Thus, questions arise: Is compactization a necessary separate step of folding or can it take place simultaneously with the native structure formation? Does nonspecific “hydrophobic collapse” facilitate folding? It is difficult to answer these questions based exclusively on the analysis of experimental data because in most cases such analysis requires certain assumptions (such as linear dependence of free energy of the unfolded state and transition state on denaturant concentration), and there may always be questions concerning the time resolution of the kinetic apparatus. The computational approach to this question seems very useful as it does not have limitations of time resolution and allows the eventual tracing of any intermediate conformation, and therefore one can be sure that nothing is “hidden” in numeric experiments. The two crucial requirements to the computational model is that model proteins should possess an astronomically large number of conformations and that simulations should reach the unique native conformation and subsequently stay in the native state, which includes native conformation and fluctuations around it. Such fluctuations are always present at finite temperature. The only available computational models in which both requirements are fulfilled are simple lattice models of proteins (Skolnick & Kolinski, 1991; Shakhnovich et al., 1991; Miller etal., 1992; Camacho & Thirumalai, 1993; Sali et al., 1994a; Socci & Onuchic, 1994). Of course, these are idealized models in which the virtue of folding is achieved by significant simplifications such as presentation of amino acid residues as structureless beads. Hence, it can tackle only such problems where detailed stereochemistry and packing of side chains may be not relevant. The most important of these are aspects of the protein folding problem concerned with formation of native chain topology, which is believed to occur during the fast stages of folding. In this study we are interested in the role of burst intermediates which may be formed at the fastest stages of folding. Therefore, we believe that the simple lattice model may be adequate to tackle this aspect of the protein folding problem. In the present work we study the role of burst intermediates using a simple lattice model of folding. We show that hydrophobic collapse is not necessary for folding. Depending on solvent conditions or amino acid composition of a protein (viz., whether monomers on average attract each other or not), burst collapse may precede folding to the native state, or compactization may proceed simultaneously with folding to the native state. In the latter case the resolution of the Levinthal paradox in our model system does not require that intermediates be formed in the process of folding.

THE MODEL We consider a protein chain positioned on a cubic lattice. An amino acid residue is presented by a structureless monomer occupying a lattice site. Residues connected by a covalent bond must occupy neighboring lattice sites. Two or more monomers cannot occupy the same site. The energy of a conformation is a sum of energies of pairwise contacts between monomers:

Biochemistry, Vol. 34, No. 9, 1995 3067

E=

C

u(EiyEj>Aij

(1)

lSi T,. This paradigm is a more consistent formulation of the intrinsic paradox in protein folding and directly implies that protein sequences must have been evolutionary selected to be able to fold, a point made by Levinthal in his original publication (Levinthal, 1969). A possible solution of this paradox was suggested recently (Shakhnovich, 1994; Abkevich et al., 1994a), where it was shown that sequences designed to have their native state as a pronounced energy minimum have kinetic access to this native conformation. This is in line with the original theoretical concepts (Bryngelson & Wolynes,

Gutin et al.

3076 Biochemistry, Vol. 34, No. 9, 1995 1989; Goldstein et al., 1992; Shakhnovich & Gutin, 1993a) since such designed nonrandom sequences have stable native structures at temperatures above T,, resolving therefore the contradiction between thermodynamics and kinetics which is a characteristic feature of random sequences.

ACKNOWLEDGMENT We are grateful to Oleg Ptitsyn for many important discussions and to Mike Morrisey for editorial help.

REFERENCES Abkevich, V. I., Gutin, A. M., & Shakhnovich, E. I. (1994a) Biochemistry 33, 10026- 1036. Abkevich, V. I.; Gutin, A. M., & Shakhnovich, E. I. (1994b) J. Chem. Phys. 101, 6052-6062. Baldwin, R. L. (1989) Trends Biochem. Sci. 14, 291-294. Bryngelson, J. D., & Wolynes, P. G. (1987) Proc. Natl. Acad. Sci. U S A . 84, 7524-7528. Bryngelson, J. D., & Wolynes, P. G. (1989) J . Phys. Chem. 93, 6902-691 5. Camacho, C. J., & Thirumalai, D. (1993a) Proc. Natl. Acad. Sci. USA. 90, 6369-6372. Camacho, C. J . , & Thirumalai, D. (1993b) Phys. Rev. Lett. 71, 2505-2508. Chan, H. S., & Dill, K. A. (1990) Proc. Natl. Acad. Sci. USA. 87, 6388. Chan, H. S., & Dill, K. A. (1994) J . Chem. Phys. 100, 92389257. Creighton, T. (1992) Proteins. Structure and Molecular Properties, W. H. Freeman & Co., New York. De Dominicis, C., Orland, H., & Lainee, F. (1985) J. Phys. Lett. 46, L463. De Gennes, P. G. (1979) Scaling Concepts in Polymer Physics, Cornel1 University Press, Ithaca, NY. De Gennes, P. G. (1985) J . Phys. Lett. 46, L639-L642. Dill, K. (1985) Biochemistry 24, 1501-1509. Dinner, A., Sali, A., Karplus, M., & Shakhnovich, E. (1994) J. Chem. Phys. 101, 1444-1451. Elove, G. A., Chaffotte, A. F., Roder, H., & Goldberg, M. E. (1992) Biochemistry 31, 6876-6883. Finkelstein, A. V., & Shakhnovich, E. I . (1989) Biopolymers 28, 1668- 1694. Finkelstein, A. V., Gutin, A. M., & Badretdinov, A. Ya. (1993) FEBS Lett. 325, 23-28. Fukugita, M., Lancaster, D., & Mitchard, M. G. (1993) Proc. Natl. Acad. Sci. USA. 90, 6365-6368. Gilmanshin, R., & Ptitsyn, 0. B. (1987) FEBS Lett. 223, 327329. Goldberg, M., Semisotnov, G., Friquet, B., Kuwajima, K., Ptitsyn, 0. B., & Sugai, S. (1990) FEBS Lett. 263, 51-56. Goldstein, R., Luthey-Schulten, Z . A., & Wolynes, P. G. (1992) Proc. Natl. Acad. Sci. U S A . 89, 9029-9033. Grosberg, A. Yu., & Shakhnovich, E. I. (1986) Sov. Phys.-JETP 64, 1284-1290. Grosberg, A. Yu., Nechaev, S. K., & Shakhnovich, E. I. (1988) J. Phys. (Paris) 49, 2095-2100. Guo, Z., & Thirumalai, D. (1994) Biopolymers (in press). Harrison, S. C., & Durbin, R. (1985) Proc. Natl. Acad. Sci. U S A . 82, 4028-4030.

Hilhorst, H. J., & Deutch, J. M. (1975) J . Chem. Phys. 63, 51535161. Jackson, S . E., & Fersht, A. (1991) Biochemistry 30, 10428-10435. Karplus, K., &Weaver, D. L. (1979) Biopolymers 18, 1421-1437. Karplus, M., & Shakhnovich, E. (1992) in Protein Folding (Creighton, T. E., Ed.) Chapter 4, W. H. Freeman and Co., New York. Khorasanizadeh, S., Peters, I. D., Butt, T. R., & Roder, H. (1993) Biochemistry 32, 7054-7063. Kim, P., & Baldwin, R. (1990) Annu. Rev. Biochem. 51, 459. Kolinski, A., Skolnick, J., & Yaris, R. (1986) J. Chem. Phys. 85, 3585-3587. Kopper, G., & Hilhorst, H . (1987) Europhys. Lett. 3, 1213-1217. Kuwajima, K., Yamaya, H., Miwa, S., Sugai, S., & Nakamura, T. (1987) FEBS Lett. 221, 115-118. Levinthal, C. (1969) in Mossbauer Spectroscopy of Biological Systems. Proceedings of a Meeting Held at Allerton House, Monticello, ZL (Debrunner, P., Tsibris, J.-C., & Munck, E., Eds.) pp 22-24, University of Illinois Press, Urbana, IL. Lifshitz, I. M., Grosberg, A. Yu., & Khohlov, A. R. (1978) Rev. Mod. Phys. 50, 683-713. Matouschek, A., Kellis, J., Jr., Serrano, L., Bycroft, M., & Fersht, A. R. (1990) Nature 346, 440-445. Miller, R., Danko, C. A., Fasolka, M. J., Balazs, A. C., Chan, H. S., & Dill, K. A. (1992) J . Chem. Phys. 96, 768-780. Miyazava, S., & Jemigan, R. L. (1985) Macromolecules 18, 534552. Orland, H., Itzykson, C., & De Dominicis, C. (1985) J . Phys. Lett. 46, L353-L355. Poland, D., & Scheraga, H. A. (1970) Theory of Helix-Coil Transitions in Biopolymers: Statistical Mechanical Theory of Order-Disorder Transitions in Biological Macromolecules, Academic Press, New York. Rey, J., & Skolnick, J. (1991) Chem. Phys. 158, 199-203. Sali, A., Shakhnovich, E. I., & Karplus, M. (1994a) J . Mol. Biol. 235, 1614-1636. Sali, A., Shakhnovich, E. I., & Karplus, M. (1994b) Nature 369, 248-251. Shakhnovich, E. I. (1994) Phys. Rev. Lett. 72, 3907-3909. Shakhnovich, E. I., & Finkelstein, A. V. (1989) Biopolymers 28, 1667- 1681. Shakhnovich, E. I., & Gutin, A. M. (1989a) Biophys. Chem. 34, 187-199. Shakhnovich, E. I., & Gutin, A. M. (198913) Europhys. Lett. 9,569574. Shakhnovich, E. I., & Gutin, A. M. (1990) Nature 346, 773-775. Shakhnovich, E. I., & Gutin, A. M. (1993a) Proc. Natl. Acad. Sci. USA. 90, 7195-7199. Shakhnovich, E. I . , & Gutin, A. M. (1993b) Protein Eng. 6,793800. Shakhnovich, E., Farztdinov, G., Gutin, A. M., & Karplus, M. (1991) Phys. Rev. Lett. 67, 1665-1668. Skolnick, J., & Kolinski, A. (1991) J . Mol. Biol. 221, 499-531. Socci, N., & Onuchic, J. (1994) J . Chem. Phys. 101, 1519-1528. Sosnick, T., Mayne, L., Hiller, R., & Englander, S. W. (1994) Nature Struct. Biol. I , 149-156. Stanley, H. E. (1971) Introduction to Phase Transitions and Critical Phenomena, Oxford University Press, New York and Oxford. Weaver, D., & Karplus, M. (1994) Protein Sci. 3, 650-668. Wetlaufer, D. B. (1973) Proc. Natl. Acad. Sci. USA. 70,697-701. B194 1997J